Interplay between FtsY and the ribosome at the SecYEG translocon

In the course of ribosome targeting to the SecYEG translocon, after the transfer of the RNC to the translocon is completed, it remains unclear whether FtsY remains bound to the translocon or leaves the complex. Previous crosslinking studies have shown that FtsY interacts with the cytosolic loops C4 and C5 of SecY, which are the main binding partners for the ribosome as well. This implies that FtsY and the ribosome compete for binding to the translocon (Kuhn et al, 2011). Initially we probed whether the ribosome can induce dissociation of FtsY from

SecYEG(ND) and confirmed that they could compete. Nevertheless, our previous data showed that FtsY binds to SecYEG(ND) at two sites via the NG and the A domains. Also that the A domain stabilizes FtsY on SecYEG(ND). Thus, there is the possibility that FtsY remains bound to SecYEG(ND), though in a different conformation, when the ribosome is bound. We tested this hypothesis and observed that FtsY and the ribosome can bind simultaneously to

SecYEG(ND) in a noncompetitive manner. Thus, confirming that FtsY participates in the ternary complex with SecYEG(ND) and the ribosome, but in somewhat different conformation.

53

Our further analysis showed that the NG domain or the A domain, both encompassing the MTS, were able to participate in the ternary complex. These data would explain the in vivo observation that the isolated NG domain which includes the MTS is able to rescue the ΔFtsY phenotype, at least partially. Provided that, the critical MTS is responsible for the localization and the isolated NG domain interacts with SRP to promote its release from the RNC (Eitan &

Bibi, 2004; Parlitz et al, 2007; Spanggord et al, 2005). Nevertheless, the A domain appeared to bind more firmly than the NG domain. Thus, confirming our previous observation that the two lipid-binding sequence assure the membrane localization and the interaction with

SecYEG(ND).

Previously, the ribosome-SecYEG interaction was quantified by surface plasmon resonance (SPR) which has shown that the complex is of high affinity and that the nascent chain further strengthens the interaction (Wu et al, 2012). By performing equilibrium titrations, monitored by FRET, we obtained similar affinities for SecYEG(ND) binding to 70S ribosomes of 50 ± 20 nM and to Lep-RNC of 9 ± 1 nM. The kinetic analysis showed that the ribosome and

SecYEG(ND) associate in a two-step mechanism, with a fast diffusion-controlled binding step, followed by a slower rearrangement step. The overall stability of the complex appeared to be low, with half-life time of the complex of 0.2 s. A stabilization of the complex in the presence of the signal sequence we did not observe in the kinetics. Most probably a difference in Kd of 5 – 7-fold is too small to be detected by kinetic measurements.

So far our data have shown that the A domain of FtsY is essential for the stabilization of FtsY on SecYEG. This effect is achieved via two sequences located at the N and the C terminus of the A domain (Braig et al, 2009). There is evidence suggesting that lipid-association stimulates the intrinsic GTPase activity of FtsY (de Leeuw et al, 2000; Lam et al, 2010). However, it is unclear how SecYEG influences the GTPase activation of FtsY and the SRP-FtsY complex. The studies that had investigated this process either used SecYEG solubilized in detergent (DDM) or only liposomes (Akopian et al, 2013b; Lam et al, 2010; Shen et al, 2012). Especially the use of DDM appears problematic, as our results show that FtsY does not interact with SecYEG in the presence of DDM. The GTPase activation of FtsY by liposomes is rather high, more than 100-fold, albeit at very high liposome concentration in the millimolar range (Lam et al, 2010(Lam et al, 2010). Considering the large available surface of the liposomes it is expected that FtsY, used

54

in 2 – 5 µM concentration, would be saturated also at micromollar concentration of liposomes.

Perhaps the use of the detergent (Nikkol) in the reactions could have perturbed the system. In our hands, FtsY is activated 2-fold both by SecYEG(ND) and ND at much lower, micromolar, yet saturating concentrations, which are also in line with the measured affinities. We further analyzed the effect on the GTPase activation of the SRP-FtsY complex by the FtsY binding partners SecYEG(ND) or ND. The GTP hydrolysis rate of the SRP-FtsY complex remained unaffected by SecYEG(ND) or ND. Also the complex of SRP with only the NG domain of FtsY exhibited the same catalytic activity as the SRP-FtsY complex. These results were in

agreement with previous observations that the NG domains of SRP and FtsY are sufficient for GTPase activation of the complex (Bahari et al, 2007; Egea et al, 2004; Eitan & Bibi, 2004; Focia et al, 2004). Taken together the data presented on the function of the N-terminal A domain of FtsY converges to the conclusion that this domain mainly stabilizes FtsY on the SecYEG, while the NG domain can bind SRP and hydrolyze GTP.

The present investigation allow for a refinement of the model of the SRP targeting pathway, by providing details of the interaction between SecYEG and FtsY and elucidating the function of the NG and A domains of FtsY (Figure 4-2). According to the current understanding of the targeting process, SRP rapidly scans the ribosomes until it is stably bound to one that has the exit tunnel filled (about 35 amino acids). As in the bacterial system there is no SRP-induced elongation arrest, continued translation may lead to the emergence of a non-signal sequence, resulting in SRP rejection (Bornemann et al, 2008; Holtkamp et al, 2012b). Alternatively, the appearance of a signal-anchor sequence (SAS) leads to a rearrangement of SRP that exposes the NG domain and strongly stabilizes the complex. This way the SRP-RNC complex is ready to be targeted to the membrane. FtsY on the other hand is localized at the membrane in the vicinity of SecYEG. There it can bind to SRP via its NG domain, which is exposed in the FtsY-translocon complex, while it remains associated to the SecYEG via the A-domain contacts.

There is evidence that following FtsY binding to SRP, the NG-NG complex translocates to the distal end of the 4.5S RNA, which is possible due to the long linker between NG and M domains of Ffh (Shen et al, 2012).

55

Figure 4-2 Model of SRP targeting pathway

SRP scans the translating ribosomes (RNC) for an emerging nascent chain. In case that the nascent chain exposes a N-terminal signal anchor sequence (SAS) of a membrane protein, SRP rearranges and binds both the ribosome and the SAS. Like that the RNC-SRP complex is ready for targeting to the membrane. FtsY is localized at the membrane in the vicinity of the SecYEG translocon. Upon interaction with SecYEG, FtsY adopts an open conformation, which allows the NG domain (yellow-orange) to interact with the NG domain of SRP (green). Due to the high affinity between the RNC and SecYEG and the increased affinity of SRP and FtsY, the complex of RNC-SRP is targeted to the SecYEG-FtsY complex to the membrane. There in a GTP-dependent manner FtsY and SRP form a complex via their NG domains. The RNC is transferred to SecYEG and SRP is released from the quaternary complex after GTP hydrolysis. FtsY remains bound to SecYEG, though in a different conformation.

Thus, upon initial targeting a quaternary complex is formed, and we speculate that this

complex is transient, because the high affinity between RNC and SecYEG will probably lead to RNC transfer to the translocon, and competition between translocon and SRP for binding to overlapping sites at the peptide exit of the ribosome. Concomitantly with the rearrangement, the GTPase activity of the complex is stimulated and the SRP – FtsY complex is released, while

56

SRP also releases the SAS. During the nascent chain translocation FtsY remains bound to SecYEG via the A domain while the NG domain remains in conformation that does not interfere with the ribosomal binding. In this state FtsY awaits the next round of targeting.

57

5. Materials and Methods